Abstract Apolipoprotein E is a class of proteins involved in lipid metabolism with functions determined by alleles of the gene APOE. The least common allele of this gene, called ?2, emerged as a putative protective variant when Schachter et al. noted an increased frequency of ?2 in French centenarians. Since then several studies have provided evidence that ?2 has a beneficial neuroprotective effect and promotes longevity and healthy aging. The New England Centenarian Study (NECS) directed by Dr. Perls has enrolled ~400 ?2 carriers since 1994, and in collaboration with the Longenity Study (LS) directed by Dr. Barzilai and other studies of centenarians produced strong evidence that ?2promotes longevity. Despite multiple research efforts however the biological mechanisms associated with ?2 are still unclear. We generated in collaboration with Novartis a proteomic dataset of ~5,000 proteins from sera of 226 NECS participants, ages 50 to 115 years, selected to be enriched of ?2 and we analyzed these data in the context of APOE genotypes. In our analysis, we discovered a preliminary set of proteins that correlate with APOE genotypes and predict different longitudinal patterns of cognitive decline. These preliminary data support the hypothesis that there are multiple plasma proteins associated with APOE genotypes that (i) can be used to predict cognitive decline or preservation better than APOE genotypes, and (ii) can inform us about potential mechanisms of action of APOE and suggest candidate interventions to prevent or delay neurodegenerative diseases and cognitive decline in aging. We propose to test this hypothesis in three ways: we will use in-depth proteomics to validate, characterize and expand the set of proteins that correlate with APOE genotypes in plasma of 50 carriers of different APOE genotypes. This analysis will likely discover additional protein biomarkers associated with APOE genotypes as well as post- translational modifications that could modify their molecular functions. We will then evaluate the effect of the expanded protein signature on aging and cognitive decline in 600 centenarians, their offspring and controls from the NECS, and replicate findings with the LS. In a subset of these subjects, we will assay protein levels in a second blood sample collected few years apart to be able to examine how changes of the protein signature predict changes of cognitive functions. We also propose to augment these data with selected markers of inflammation and free fatty acids that are important factors in cognitive aging and conduct integrative analyses of the data collected in the various aims to model hypothetical mechanisms linking APOE genotypes to cognitive aging. In summary, we have assembled a competent, interdisciplinary team of investigators to thoroughly evaluate and characterize protein signatures of APOE genotypes that could become multipurpose, potent biomarkers of cognitive functions change and probable AD. The results of this work will suggest possible mechanisms that determine the neuroprotective effect of the APOE ?2 allele.